Trainability barriers and opportunities in quantum generative modeling

Abstract

Quantum generative models provide inherently efficient sampling strategies and thus show promise for achieving an advantage using quantum hardware. In this work, we investigate the barriers to the trainability of quantum generative models posed by barren plateaus and exponential loss concentration. We explore the interplay between explicit and implicit models and losses, and show that using quantum generative models with explicit losses such as the KL divergence leads to a new flavour of barren plateaus. In contrast, the implicit Maximum Mean Discrepancy loss can be viewed as the expectation value of an observable that is either low-bodied and provably trainable, or global and untrainable depending on the choice of kernel. In parallel, we find that solely low-bodied implicit losses cannot in general distinguish high-order correlations in the target data, while some quantum loss estimation strategies can. We validate our findings by comparing different loss functions for modelling data from High-Energy-Physics.

21+44 pages, 10+2 figures

References in corpus (60)

• Quantum Computing in the NISQ era and beyond
• Quantum Machine Learning
• Quantum algorithm for solving linear systems of equations
• Barren plateaus in quantum neural network training landscapes
• Quantum principal component analysis
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• Evaluating analytic gradients on quantum hardware
• Quantum Computational Supremacy
• Connecting ansatz expressibility to gradient magnitudes and barren plateaus
• Quantum generative adversarial learning
• Quantum advantage in learning from experiments
• Quantum singular value transformation and beyond: exponential improvements for quantum matrix arithmetics
• Quantum Generative Adversarial Networks for Learning and Loading Random Distributions
• An initialization strategy for addressing barren plateaus in parametrized quantum circuits
• A generative modeling approach for benchmarking and training shallow quantum circuits
• Quantum-assisted quantum compiling
• Absence of Barren Plateaus in Quantum Convolutional Neural Networks
• Differentiable Learning of Quantum Circuit Born Machine
• Opportunities and challenges for quantum-assisted machine learning in near-term quantum computers
• Effect of barren plateaus on gradient-free optimization
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• Training of Quantum Circuits on a Hybrid Quantum Computer
• Stochastic gradient descent for hybrid quantum-classical optimization
• Quantum machine learning beyond kernel methods
• Exploiting symmetry in variational quantum machine learning
• The SWAP test and the Hong-Ou-Mandel effect are equivalent
• Trainability of Dissipative Perceptron-Based Quantum Neural Networks
• Quantum generative adversarial network for generating discrete distribution
• Group-Invariant Quantum Machine Learning
• The Born Supremacy: Quantum Advantage and Training of an Ising Born Machine
• Equivalence of quantum barren plateaus to cost concentration and narrow gorges
• Entanglement Devised Barren Plateau Mitigation
• Barren plateaus preclude learning scramblers
• Higher Order Derivatives of Quantum Neural Networks with Barren Plateaus
• Tree Tensor Networks for Generative Modeling
• Information Perspective to Probabilistic Modeling: Boltzmann Machines versus Born Machines
• Theory for Equivariant Quantum Neural Networks
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• On the Quantum versus Classical Learnability of Discrete Distributions
• Analyzing the barren plateau phenomenon in training quantum neural networks with the ZX-calculus
• Out-of-distribution generalization for learning quantum dynamics
• On the practical usefulness of the Hardware Efficient Ansatz
• Synergy Between Quantum Circuits and Tensor Networks: Short-cutting the Race to Practical Quantum Advantage
• Generative model benchmarks for superconducting qubits
• Variational inference with a quantum computer
• Style-based quantum generative adversarial networks for Monte Carlo events
• Can Error Mitigation Improve Trainability of Noisy Variational Quantum Algorithms?
• Expressibility and trainability of parameterized analog quantum systems for machine learning applications
• Mitigating Barren Plateaus with Transfer-learning-inspired Parameter Initializations
• Dynamical simulation via quantum machine learning with provable generalization
• Interpretable Quantum Advantage in Neural Sequence Learning
• Generative machine learning with tensor networks: benchmarks on near-term quantum computers
• On the Sample Complexity of Quantum Boltzmann Machine Learning
• Tight and Efficient Gradient Bounds for Parameterized Quantum Circuits
• Unsupervised Quantum Circuit Learning in High Energy Physics
• Conditional Born machine for Monte Carlo event generation
• Introducing Non-Linear Activations into Quantum Generative Models
• A Continuous Variable Born Machine
• F-Divergences and Cost Function Locality in Generative Modelling with Quantum Circuits
• Generative modeling with projected entangled-pair states

Cited by in corpus (18)

• Barren Plateaus in Variational Quantum Computing
• Does provable absence of barren plateaus imply classical simulability?
• On fundamental aspects of quantum extreme learning machines
• Variational quantum simulation: a case study for understanding warm starts
• Quantum Convolutional Neural Networks are Effectively Classically Simulable
• Opportunities and challenges of quantum computing for climate modelling
• Symmetry-invariant quantum machine learning force fields
• Variational measurement-based quantum computation for generative modeling
• Error-mitigated photonic quantum circuit Born machine
• QSlack: A slack-variable approach for variational quantum semi-definite programming
• Efficient quantum-enhanced classical simulation for patches of quantum landscapes
• Machine-Learning Insights into the Entanglement-trainability Correlation of Parameterized Quantum Circuits
• Cross-platform hardware benchmark of style-based quantum GANs for data augmentation on superconducting and trapped-ion processors
• Pitfalls when tackling the exponential concentration of parameterized quantum models
• Variational Inference for Quantum HyperNetworks
• Expressive equivalence of classical and quantum restricted Boltzmann machines
• Quantum Chebyshev Probabilistic Models for Fragmentation Functions
• Quantum Machine Learning for State Tomography Using Classical Data